RESEARCH OPPORTUNITIES IN IMAGE ANALYSIS/COMPUTER VISION

Melbourne Australia.

A number of opportunities exist for suitable candidates at Postdoctoral Level or PhD student level.

These include:
Postdoctoral Position in Video Tracking (two year appointment from mid-2008 - although there may be the possibility of an extension)
Postdoctoral Position in Robust Statistical Methods of Model Fitting (up to four year appointment from early 2008)
Several PhD/Masters Scholarships - associated with the above projects + one scholarship in Biomedical Image Analysis (two - Masters - to three years - PhD - from early 2008)

These positions will be filled as soon as suitable applicants can be found. Interested applicants should send their details as soon as possible to the contact persons listed below.

For Further Details on each project - see the details below.

***********************************************
PhD Scholarship in Brain Structure Mapping
Monash University (Melbourne Australia) - Join Project led by
Prof. David Reutens, Monash University Faculy of Medicine
Dr. Richard Beare, Monash University Faculy of Medicine
Prof. David Suter, Monash University Dept. of Elect.& Comp. Systems Engineering
Send Expressions of Interest to Dr. Richard Beare Richard.Beare@med.monash.edu.au WITH A CC to Prof. David Suter d.suter@eng.monash.edu.au

A major challenge in neuroscience research is detecting and characterizing changes in brain structure. A PhD scholarship is available in the Faculty of Medicine and the Faculty of Engineering at Monash University to develop novel image analysis and statistical techniques for examining the structure of the brain of humans and animal models. This work will contribute to the success of leading edge neuroscience initiatives such as the Australian Mouse Brain Mapping Consortium.

Candidates should have strong mathematical or statistical skills and experience in image analysis and software development would be beneficial.

***********************************************

Postoctoral Position and PhD Scholarship in Visual Tracking (A stochastic geometrical approach)
Melbourne/Monash Universities (Melbourne, Australia) - Project led by:
A/Prof. Ba-Ngu Vo, Melbourne University Dept. Elect. Engineering
Prof. David Suter, , Monash University Dept. of Elect.& Comp. Systems Engineering
Send Expressions of Interest to Assoc. Prof. Vo bv@ee.unimelb.edu.au WITH A CC to Prof. David Suter d.suter@eng.monash.edu.au

This project combines Bayesian filtering theory with stochastic geometry to address the problem of tracking multiple-object tracking in video sequences. The Bayesian framework is employed for its suitability to on-line implementation and its adaptability to the fusion of heterogeneous sources of information. The stochastic geometric modelling provides a solid foundation for the systematic study of the problem and allows principled approximations to made. Stochastic geometrical models, including deformable templates and point processes have long been used by statisticians to develop techniques for object recognition in static images [24]. However, their use has been largely overlooked in the tracking area until recently. In tracking with point measurements, an approximation based on point process theory known as the Probability Hypothesis Density (PHD) filter has been shown to alleviate the data association problem and remove the bulk of the computational load [25-28]. In this project we will develop a PHD-type filter for video data to arrive at an efficient algorithm for tracking multiple objects.
[24] A. J. Baddeley and M. N. M. v. Lieshout, "ICM for object recognition," in Computational Statistics. vol. 2, Y. D. a. J. Whittaker, Ed. Heidelberg-New York: Springer, 1992, pp. 271-286.
[25] R. Mahler, "Multi-target Bayes filtering via first-order multi-target moments," IEEE Trans. Aerospace and Electronic Systems, vol. 39, pp. 1152–1178, 2003.
[26] B. Vo, S. Singh, and A. Doucet, "Sequential Monte Carlo methods for Bayesian Multi-target filtering with Random Finite Sets," IEEE Trans. Aerospace and Electronic Systems, vol. 41, pp. 1224-1245, 2005.
[27] B. Vo and W. K. Ma, "The Gaussian mixture Probability Hypothesis Density filter," IEEE Trans. Signal Processing, vol. 54, pp. 4091-4104, 2006.
[28] R. Mahler, Statistical Multisource-Multitarget Information Fusion: Artech House, 2007.

Candidates should have strong mathematical or statistical skills and experience in image analysis and software development would be beneficial.

*********************************************
Postoctoral Position and PhD Scholarship in Robust Statistical Fitting
Swinburne/Monash Universities (Melbourne, Australia) - Project led by:
Prof. David Suter, , Monash University Dept. of Elect.& Comp. Systems Engineering
A/Prof. Ali Bab-Hadiashar, Swinburne University Faculty of Engineering and Industrial Sciences

Send Expressions of Interest to Assoc. Prof. Bab-Hadiashar abab-hadiashar@swin.edu.au WITH A CC to Prof. David Suter d.suter@eng.monash.edu.au

Our aim is to study the essential (and ubiquitous) problem of automatically segmenting visual data into meaningful parts - a form of model fitting: but in the face of multiple objects, missing and noisy data, large data volumes, and unknown form (model) and number of objects. These latter characteristics lift the problem beyond the reach of standard statistical fitting approaches.
There have been studies concentrating on optimal fitting (Kanatani 1996) (ignoring such things as outliers, multiple structures, data sample density etc.), or on robust fitting in the presence of multiple structures ((Meer 2004) not only provides a good overview if these approaches but also a good discussion of why traditional robust statistics is inadequate for such a setting) but generally ignoring small sample issues, model selection, and a plethora of other practical issues. This project is an ambitious attempt to provide a more complete theory and practical methodology.
The project will focus on range data (i.e., laser scan and 3D reconstruction from images) and motion estimation/segmentation. It will build upon the work of the team (see references below).
(Schindler 2006) K. Schindler, J. U, and H. Wang. Perspective n-view Multibody Structure-and-Motion through Model Selection. Proc. 9th European Conference on Computer Vision, Graz, Austria, 2006.
N. Gheissari, A. Bab-Hadiashar, and D. Suter. Parametric model-based motion segmentation using surface selection criterion. Computer Vision and Image Understanding, 102(2):214-226, 2006.
K. Schindler and D. Suter. Two-view multibody structure-and-motion with outliers through model selection. IEEE Trans. Pattern Analysis and Machine Intelligence, 28(6):983-995, 2006.
P. Chen and D. Suter. An analysis of linear subspace approaches for computer vision and pattern recognition. International Journal of Computer Vision, 68(1):83-106, 2006.
H. Wang and D. Suter. Robust Adaptive-Scale Parametric Model Estimation for Computer Vision. IEEE Trans. Pattern Analysis and Machine Intelligence, 26(11):1459–1479, November 2004.
P. Chen and D. Suter. Recovering the missing components in a large noisy low-rank matrix: Application to SFM. IEEE Trans. Pattern Analysis and Machine Intelligence, 26(8):1051– 1063, August 2004.
H. Wang and D. Suter. MDPE: A very robust estimator for model fitting and range image segmentation. Int. J. of Computer Vision, 59(2):139–166, September 2004.
R. Hoseinnezhad, A. Bab-Hadiashar, and D. Suter. Finite sample bias of robust scale estimators in computer vision problems. In Lecture Notes in Computer Science, International Symposium on Visual Computing (ISVC06), volume 4291, pages 445-454, Heidelberg, 2006. Springer-Verlag.

(Kanatani 1996) K. Kanatani, Statistical Optimization for Geometric Computation: Theory and Practice, Elsevier Science, Amsterdam, The Netherlands, 1996
(Meer 2004) P. Meer: Robust techniques for computer vision. Emerging Topics in Computer Vision, G. Medioni and S. B. Kang (Eds.), Prentice Hall, 107-190, 2004.


Candidates should have strong mathematical or statistical skills and experience in image analysis and software development would be beneficial.